Brain Region Specific Transcriptome Association Analysis For Alzheimer's Disease Neuropathology

Abstract

Objective: Alzheimer’s disease (AD), the leading cause of dementia, is a heterogeneous neurodegenerative in terms of genetics, clinical presentations, and the density and distribution of neuropathologic lesions. Our research aimed to explore the relationships between transcriptome profiles and the levels of three neuropathological markers - amyloid-β, p-tau, and pTDP43 pathology - as determined by quantitative immunohistochemistry. These markers were studied in brain region samples from the Aging, Dementia, and Traumatic Brain Injury Study. Additionally, we sought to understand how the expression of certain genes, regulated genetically, is associated with various neurodegenerative diseases. Method: We conducted a quantitative neuropathology analysis for three neuropathological markers - amyloid-β, p-tau and pTDP43 pathology, in 377 brain samples taken from four brain regions of 94 subjects enrolled in the Adult Changes in Thought (ACT) study using Pearson correlation and hierarchical clustering analysis. Then, a transcriptome-pathology association analysis was performed for 19,271 protein-coding genes. To account the correlations between data from the same brain region, a mixed linear regression model was applied. We also conducted an association analysis between predicted genetically regulated gene expression with various neurodegenerative diseases. This was performed using the TWAS method FUSION, focusing on the genes identified through initial transcriptome-pathology association analysis. Results: We found that the correlation coefficient between amyloid-β and p-tau ranged from 0.43 to 0.63, indicating a moderate positive correlation. Meanwhile, the correlation between pTDP-43 and p-tau was between 0.08 to 0.33, and between pTDP-43 and amyloid-β, it ranged from -0.06 to 0.14. These findings suggest that pTDP-43 pathology was largely independent from amyloid-β but exhibits a weak correlation with p-tau. The clustering analysis discovered a subgroup of AD patients characterized by a significantly high level of pTDP-43 pathology, while not showing elevated levels of more commonly studied amyloid-β and p-tau pathology. The analysis of transcriptome-pathology association identified 3, 11 and 9 genes associated with amyloid-β, pTau and pTDP-43 pathologies, respectively, reaching genome-wide significant level. The analysis of pathway enrichment within the transcriptomic profiles linked to various neuropathological conditions uncovered specific molecular pathways associated with distinct pathology measures, notably for pTDP-43. Finally, our research identified that the predicted genetic regulated expression levels of three genes, which were associated with pTDP-43 pathology, are linked to AD, ALS and FTD.